See the northern lights before the sky decides.

Think of Kp as the sky’s “storm strength” meter. Low Kp usually means auroras stay near the poles; higher values can bring them into more southern skies.

Kp is a 3-hour geomagnetic index derived from mid-latitude magnetometer disturbances on a quasi-logarithmic scale from 0 to 9. It summarizes how strongly Earth’s magnetic field is being pushed by solar wind energy.

We use it as the broad “how active is the magnetosphere?” input, then blend in darkness, clouds, and Bz so the forecast is more location-aware than raw Kp alone.

Bz is the “direction” of the solar wind’s magnetic field. Negative Bz is generally good for auroras because it opens the door for energy to enter Earth’s system.

Bz is the north–south component of the interplanetary magnetic field (IMF) in geocentric solar magnetospheric coordinates. A sustained southward orientation (negative values) favors magnetic reconnection at the dayside magnetopause.

Short negative spikes are less useful than a steady southward period, so the forecast weighs Bz alongside solar wind speed and darkness rather than treating it as a yes/no flag.

Clouds are the curtain in front of the show. Even a strong aurora won’t matter if the view is covered overhead.

We use an estimated cloud fraction to approximate optical obstruction. It’s not the same as visibility or transparency, but it’s a strong first-order blocker for seeing aurora from the ground.

The forecast combines cloud estimate with the actual observing window, because a clear sky for one hour can beat a mediocre all-night percentage.

This is the part of the night when the sky is truly dark enough for aurora to stand out.

We approximate the window using local solar geometry, looking for the interval between astronomical twilight boundaries and the darkest part of the night.

A decent aurora under a bright sky can look invisible, so we treat darkness as a separate ingredient from magnetic activity and cloud cover.